1. Field of the Invention
The present invention relates to an image processing method and an apparatus therefor, whereby additional information is superimposed on a captured image.
2. Description of the Related Art
Video cameras and digital cameras for use in capturing moving images have been widely spread and used for example for photographing children in an amusement park or the like. These video cameras and digital cameras are provided with an imaging section, a recording section for recording acquired image data, and a display device, such as an LCD, for displaying images which are being currently captured or images which have been recorded therein. The video cameras and digital cameras are mostly equipped with function of taking still images as well as moving images.
Recently, such a video camera or digital camera has been known that is equipped with an augmented reality representing function, whereby a variety of additional information or augmented reality (AR) information may be superimposed on an image of a physical space, e.g. an image of a landscape, which is captured and displayed by the camera. A system described in Japanese Patent Laid-open Application Publication No. 2010-039556 predicts the time of completion at which a CG generating section completes generating a virtual space image to be superimposed on the physical space image, and then predicts the position and orientation of a viewpoint at the predicted generation complete time. The CG generating section generates the virtual space image in a condition as visible from the predicted position and orientation of the viewpoint, and superimposes the same on the physical space image. Thus, the virtual space image may be composited on the physical space image without any unnatural impression, even while it takes time to generate the virtual space image.
In the system described in Japanese Patent Laid-open Application Publication No. 2010-039556, a virtual object is retrieved from a database on the basis of information on the detected position and orientation. If the accuracy of the information on the detected position and orientation is low or the detected position and orientation deviate from correct ones, the virtual object will be retrieved on the basis of the incorrect position and orientation instead of the correct ones. As a result, it sometimes happens that the superimposed virtual space image does not match the physical space image.
A terminal device described in Japanese Patent Laid-open Application Publication No. 2008-193640 reads a digital code, such as QR code (a trademark), as the code being represented on a captured image, and acquires additional information on the basis of the read digital code. Then, the acquired additional information is overlaid on the captured image as displayed on a display means. Thus, only such additional information that corresponds to the digital code represented on the captured image is overlaid on the captured image, but such additional information that is unsuitable for the captured image will not be overlaid thereon.
In order to make sure to acquire additional information from a captured image, the captured image is required to represent the digital code with little blur. In contrast, for the sake of fluent or fluid representation of successive moving images, the subjects other than the digital code should be moderately blurred in the captured images. That is, images suitable for reliable acquisition of additional information discord with images suitable for fluent representation of moving images. Therefore, in the device in which additional information is acquired from those images which are captured to be displayed on the display means, as described in Japanese Patent Laid-open Application Publication No. 2008-193640, it is hard to achieve reliable acquisition of additional information and fluid representation of moving images at the same time.